On dÿtus, ancient and modern

Robert Blust On datus, ancient and modern University of Hawai‘i

Most entries in the 877-page Maranao dictionary (McKaughan & Macaraya 1967) occupy a single line, and some are glossed with a single word. One notable exception to this pattern is the entry dÿto?, which occupies eleven lines, and three more if larger collocations that contain this base are included:

dÿto? ‘chieftain, leader, boss, lord, master, king, gentleman’ 1. Sekÿniÿnidÿto? ko mÿito? ÿ iNed. ‘He is the chieftain of the village.’ 2. Sekÿniÿnidÿto? omÿNÿ gomÿgÿlebek. ‘He is the leader of the laborers.’ 3. Sekÿniÿnidÿto? ko opisinÿ mi. ‘He is the boss in our office.’ 4. Sekÿniÿnidÿto? o bisÿiÿ?. ‘He is the master of the slave.’ 5. Sekÿniÿnidÿto? sÿÿloNÿn ÿ mÿori. ‘He is the king of

* It is a special privilege for me to offer this paper to Howard McKaughan, who in many ways is the reason I became a linguist. During the 1960s I was a lowly private stationed at Schofield Barracks in Wahiawã, Hawai‘i. Like others in my position, I had eight hours a day to check a three-quarter-ton truck to make sure there was water in the radiator and air in the tires. When this ceased to be a challenge, I seized the first opportunity to enter a newly opened language school, where I was able to study Indonesian and visit the Republic of near the end of the Sukarno period. From this point I knew that I wanted to do something with my newfound language skill, but I wasn’t sure what. Then one evening a fellow language schooler told me that a University of Hawai‘i professor was giving a lecture on base about linguistics. This sounded interesting, and I dropped by somewhere in the middle. There at the front of the room was a large, imposing, and energetic speaker. It was Howard McKaughan, aged forty-three. He told us how important linguistics was, how difficult it could be, and how much still needed to be done. After the talk I worked up the courage to approach him, and ask how I might study linguistics myself. He suggested that if I took a local discharge I should come and see him after my tour of duty was over. I nearly missed that chance, as my unit was mobilized for Vietnam less than a month before my discharge date. Since I had under ninety days of duty left, I stayed behind, but many of my friends went. Some never returned, and some are memorialized by name on a certain wall in Washington, DC. When I was free at last I enrolled at the University of Hawai‘i, and made an appointment to see Howard. He advised me on my study program, and one part of that conversation was especially salient: “Don’t try to do everything,” he said. Ever since then I have made a concerted effort not to do everything, although perhaps with only mixed success. Howard had founded the Department of Linguistics at the University of Hawai‘i, but became dean of the Graduate Division shortly after I was admitted to the program, and as a result I never had a chance to take a course from him. I did, however, purchase a copy of the McKaughan & Macaraya Maranao dictionary, which was published about the time of my arrival. Within three years I had worn it out, falling to pieces. When I mentioned this to Howard he was so pleased that he gave me a new copy, one that I still have today. One of my fellow students who studied referred to this dictionary as “the phone book”—a commentary on its great bulk and one-word definitions for many entries. In tribute to Howard, and what he did to change my life so many years ago, I want to dedicate this paper to one entry in that dictionary that departs from this stereotyped description.

Loren Billings & Nelleke Goudswaard (eds.), Piakandatu ami Dr. Howard P. McKaughan, 36–51. Manila: Linguistic Society of the and SIL Philippines, 2010. ON DATUS, ANCIENT AND MODERN 37

the world to come.’ 6. Apiÿÿndÿ sekÿniÿnmÿtÿgÿo? nÿ dÿto? den. ‘Wherever he is he is a gentleman.’ dÿto? ÿÿdil ‘title of nobility’ dÿto? ÿ kÿbogÿtÿn ‘title of nobility’ dÿto? imÿm ‘title of religious leader’

One can conclude from this greater-than-average degree of attention that the word dÿto? describes an important notion in Maranao society. LeBar (1975:23), describing the Muslim populations of the southern Philippines as a whole, notes: Perhaps the most characteristic feature of Moro society, apart from religion, has been the so-called “dÿto system,” whereby all persons in a district considered themselves allied to a local dÿto (dÿtu), sultan, or similarly titled individual. Under this system the Moros were organized into numerous petty states or principalities, in a manner not unlike the early coastal Malay sultanates of Sumatra, Java, Borneo, and Celebes. Leadership was largely restricted to members of aristocratic family lines, whose claim to authority rested ultimately on Koranic scripture; actual power was a matter of individual charisma backed by the number and strength of one’s entourage. This definition suggests a strong association between the notion of the dÿtu and the adoption of Islam. However, there are reasons to believe that the dÿtu system, whatever its original character, existed long before Islam was introduced to the southern Philippines or other parts of insular Southeast Asia. Dempwolff (1938) posited “Uraustronesisch” *datu ‘Sippenhaupt’, hence ‘head of a kin group’.1 As evidence he cited Tagalog dÿtoÛ ? ‘high priest’ (Oberpriester), Toba Batak dÿtu ‘sorcerer, shaman’ (Zauberpriester), Javanese rÿtu ‘prince’ (Fürst), Malay dÿtuk ‘head of a kin group’ (Sippenhaupt), Fijian rÿtu ‘master, owner; Sir’ (Herr), and Samoan lÿtu ‘one in charge of construction’ (Baumeister). Despite problems to be noted below which force us to abandon the last two members of this cognate set, these forms show a considerably wider range of meanings than are associated with reflexes of *datu in the Muslim groups of the southern Philippines, and naturally raise questions about what this term might originally have meant, and what change paths it might have followed in producing the attested semantic diversity of its reflexes. Dempwolff’s tentative hypothesis was that the *datu was the head of a kin group, but the justification for this semantic inference is by no means clear from the supporting evidence he cites, as only Malay dÿtuk is given with a meaning at all close to that of the reconstructed form. The gloss that he gives for this term may have been inspired by a consideration of a wider range of evidence than he was able to incorporate in the Vergleichende Lautlehre, or it may simply have been an expedient subject to change, since he had not worked out a systematic account of early Austronesian (AN) kinship, or social organization. Yet, as will be shown, it comes close to what a more systematic examination of the evidence appears to justify. To pursue this question further, we need to identify reflexes of *datu in many more AN languages, and to collect the fullest possible information about the meaning of these terms. A natural way to do this is to follow the map, as it were, beginning in Taiwan, then proceeding southward through the Philippines from the Batanes Islands to the Sulu Archipelago, on to

1 Data in this paper preceded by an asterisk are reconstructed historical forms. Most transcriptions follow IPA notation. A notable exception is y for the palatal glide and j (sometimes dj)fora post-alveolar affricate. Occasionally letters are capitalized following the usage in the original source. 38 ROBERT BLUST

Borneo and the rest of western Indonesia, and then eastward through Indonesia into the Pacific. In doing this, the first thing that strikes the investigator is the total absence of such terms in Taiwan. Since the Formosan aboriginal languages apparently have no reflexes of *datu, we must conclude that, so far as the available evidence permits us to infer, this word can be assigned only to Proto-Malayo-Polynesian (PMP), the hypothetical common ancestor of all non-Formosan AN languages, but not to PAN itself. Although the term turns up in the northern Philippines, it appears to occur only as a loanword, with reference to Muslim populations in the far south:

Ilokano dÿtoÛ ‘Muslim ruler’ (Rubino 2000) Pangasinan dÿtuÛ ‘leader of a tribal group, chief’ (Benton 1971) Kapampangan dÿtuÛ ? ‘leader, rarely used except with reference to southern Philippines or to olden days’ (Forman 1971; not in Bergaño 1860)

In the central Philippines the term is used in ways which more clearly suggest that it was applied to the local societies prior to the introduction of Christianity:

Tagalog dÿtoÛ ? ‘high priest’ (Laktaw 1914); ‘tribal chief in pre-Christian days’ (Panganiban 1966) Bikol dÿtoÛ ? ‘(archaic) headman, chief; also referring to the rich and influential members of pre-Hispanic Bikol society’ (Mintz 2004) Aklanon dÿtuÛ ? ‘datu, ruler, king; rich, wealthy, powerful (through money)’ (Salas Reyes et al. 1969) Masbatenyo dÿtoÛ ? ‘(from Cebuano) chief, ruler’ (Wolfenden 2001) Cebuano dÿtuÛ ? ‘rich, wealthy; title of a chief, now said only to Muslim leaders’ (Wolff 1972)

It is noteworthy that reflexes of *datu in the northern and central Philippines are confined almost exclusively to lowland populations; interior groups such as the Isnag, Itawis, Bontok, Kankanaey, Ifugao, or Hanunóo simply lack the word. This distribution implies two things. First, it suggests that the word in the central and northern Philippines is a loan. Second, it suggests that the word was associated with the spread of Islamic culture, a development that had begun to affect the central Philippines in pre-Hispanic times, but was effectively terminated outside and Sulu when the Spanish began the Christianization of the lowland Philippines. This brings us to the southern Philippines, where the word still applies to many contemporary cultures, including the following:

Maranao dÿto? ‘chieftain, leader, boss, lord, master, king, gentleman’ (McKaughan & Macaraya 1967) Tausug dÿtu? ‘a prince, any male member of a royal family, datu’ (Hassan et al. 1994) Mapun dÿtu? ‘chief; headman; any male member or descendant of a chief’s family’; ÿnÿkdÿtu? ‘prince; the child of a chief’ (Collins et al. 2001) Yakan dÿtu? ‘chief; headman; datu, a rank bestowed on certain people; Yakans didn’t have datus originally’ (Behrens 2002) Mansaka dÿto? ‘leader, chief; a dÿto? is usually someone who has wealth’ (Svelmoe & Svelmoe 1990) ON DATUS, ANCIENT AND MODERN 39

Western Bukidnon Manobo dÿtu? ‘chief; leader’ (Elkins 1968) Binukid dÿtu? ‘chieftain; mediator’ (Post & Gardner 1992) Tboli dÿtu? ‘a traditional leader over a group of people living in a particular geographical area [...] Each area had its own leader, and each leader had equal authority, with no one leader over the others. It was expected that the position would be passed on from father to son. The leader was responsible for the welfare of those under his authority [...] His main function was to settle disputes which required traditional items to help pay the fines of those unable to pay. In return the man helped in this way was under obligation to work for the leader until he had paid his debt [...] If it were a woman who had been helped, the leader either married her himself or he had her married to someone else so that through the dowry he could get back the property he had used on her behalf. [...]’ (Awed et al. 2004) Tiruray dÿtu? ‘a Moslem nobleman; to trade mountain forest products to a coastal Maguindanao dÿtu? in return for iron tools and other items; a male leader among superhuman beings’ (Schlegel 1971)

In addition to these dictionary entries, discussions of the sociocultural role of the dÿtu appear in a number of anthropological publications. Cole (1913), describes the dÿtu as centrally important in the social organization of the animist Bagobo of Mindanao, but adds (1913:52) that despite little evidence for intermarriage, trade between the coastal Muslims and interior animists was sufficiently important “for the Moro to exert a marked influence on the religious and civil life of the wilder tribe, and to cause them to incorporate into their language many new words and terms.” This suggests that the term dÿtu in Bagobo probably is a loan from one or more of the numerically and culturally dominant Muslim populations of the coast. In a similar vein, LeBar (1975:34), describing the Subanun of the Zamboanga peninsula, notes, “Much of Zamboanga was formerly claimed by the sultans of Magindanao, whose agents, Muslim dÿtus in the coastal settlements, extracted tribute from the inland Subanun.” Reflexes of *datu in the languages of Borneo also show a decided preference for coastal languages that are spoken by Muslim populations. The word is unknown in interior languages such as Kayan, Kenyah, Kelabit, Maloh, Bödayuh, or even Iban, except as a recent loan from Malay. It is, however, reported as a kinship term in most of the Barito languages of southeast Borneo: Lawangan, Dusun Malang, Dusun Witu, Paku, Maanyan, Kapuas, Ba’amang, Murung dÿtu?, Dusun Deyah, Samihim dÿtu ‘PaPaPa’, hence ‘great grandparents’ (Hudson 1967). Curiously, Hudson gives a different sense for ‘great-great grandparents’ (‘PaPaPaPa’), which is marked either by a reflex of *muyaN (Lawangan, Dusun Malang, Paku, Maanyan, Kapuas, Ba’amang), by another term, or by no special word. The reflex of *datu in these languages thus correlates with a specific generation, and cannot be given the generalized gloss ‘ancestor’. It would be desirable to have fuller information, but what we do have reinforces the impression that reflexes of *datu have been borrowed throughout the Philippines and in most parts of Borneo in connection with Islamic cultural institutions. As will be seen, the form of this word outside the Philippines generally points to *datu, and the final glottal stop thus marks it as a likely borrowing of Malay dÿto?, dÿtok which Wilkinson (1959) glossesasfollows.

‘head of family; elder [...] Commonly shortened to? and now used 1. as a term of relationship to or of a grandfather [...] 2. as a term of distinction to or of any great non-royal chief [...] 3. in the compound dÿto? perempuÿn (to? puÿn)of 40 ROBERT BLUST

the wife of a major Chief [...] 4. in the compound dÿto? mudÿ (to? mudÿ)of the heir of a Chief [...] 5. as a term of respect to any man of age and standing’ (to which he adds dÿto? nenek ‘ancestors generally’)

Wilkinson also gives two other forms of this word:

dÿtoN ‘grandfather; my grandfather; = dÿto?, but very literary and used only of (or to) one’s own grandfather’ dÿtu ‘ruler; chief [...] This is the old Malay sovereign-title borne by the Sri Vijaya king of the Kota Kapur inscription (AS 608 = AD 686)’ Kedÿtuÿn ‘kingdom (of Sri Vijaya)’

These variants are important, as they show 1. that dÿtuk (i.e., [datu?]), like dÿtoN, probably contains a vocative suffix, and 2. that a reflex of *datu was used in Malay before the introduction of Islam, but apparently became an Islamic title in the specific form dÿtuk.2 Wilkinson also cross-references dÿtu to rÿtu, a Javanese form borrowed back into Malay, and used in the modern period primarily in the sense of ‘queen’. Given the epigraphic evidence that a reflex of *datu designated the sovereign of the Malay-speaking Hindu-Buddhist state of Srÿvijaya in the late seventh century, it seems likely that the related terms in Philippine languages reflect post-Islamic developments in Malay, and were introduced to Sulu and Mindanao by Brunei missionaries, spreading from there to the central and northern Philippines. What, then, did *datu mean before its reflexes were applied to social institutions introduced from other parts of the world (notions of kingship that were brought in with Indianization, notions of loyalty to the sultan that replaced these with the introduction of Islam)? To penetrate further back in time, we must examine reflexes of *datu in more languages, especially (although not exclusively) in languages that have been less subject to strong contact influence from Malay. The following forms are known from other languages of western Indonesia:

Acehnese dÿtu or dÿto? (sometimes shortened to to?) ‘grandfather, ancestor; also term of address for elders; title for the guardians of traditional law on the west coast; title of animals or things that are regarded as holy or especially feared’ (Kreemer 1931) Gayo dÿtu ‘ancestors in the female line’ (Melalatoa et al. 1985) Karo Batak dÿtuk ‘title of the head of the urungs (federation of different villages related through a “mother” village); he is subordinate to the sultan; title of powerful magicians, spirits and deities’ (Neumann 1951) Toba Batak dÿtu ‘shaman, magician, soothsayer, doctor who makes offerings to the spirits, reads the signs in chicken entrails, selects auspicious days, and whose advice is sought in all matters of daily concern’ (Warneck 1977); hÿ-dÿtu-on ‘skill in the science of the dÿtu, which could sometimes be found in a rÿdjÿ’ (Vergouwen 1964:132)

2 See Blust (1979) for evidence that *-q, *-N and shortening of disyllables were used as alternative means of forming the vocative of kin terms in PMP, and that *-q ‘vocative’ was reanalyzed as -k. Other examples in Malay are kÿkÿk ‘elder sister; (sometimes) elder brother’, kÿkÿN ‘my elder brother or sister’ (< PMP *kaka ‘elder sibling of the same sex’) and ÿdik ‘younger sibling’, ÿdiN ‘younger sibling’ (< *adi < *wadi < PMP *huaji ‘younger sibling of same sex’). Wilkinson (1959) cites both kÿkÿN and ÿdiN as vocative forms confined to Malayo-Javanic romance. ON DATUS, ANCIENT AND MODERN 41

Dairi-Pakpak Batak dÿtu ‘function among animists (‘orang kafir’), worshipping idols; shaman’ (Manik 1977) Simalur dÿtu? ‘head of a kin group and guardian of traditional law’ (Kähler 1961) Minangkabau dÿtu ‘title for a native medical practitioner; clever, learned’; dÿtu-ÿ? ‘title for administrators of customary law; term of address to an elder; grandfather’ (van der Toorn 1891) Sundanese rÿtu ‘prince, king, princess; in South Bantam the highest title for noblewomen’; a loan from Javanese (Coolsma 1930) Old Javanese rÿtu ‘king (rarely queen)’; kÿ-rÿtu-n ‘kingship, royal dignity’ (Zoetmulder 1982) Javanese rÿtu ‘king, queen, monarch’, krÿton (< *ka-ratu-an) ‘palace, court’ (Horne 1974) Madurese rÿto ‘king’, kÿ-rÿto-n ‘palace’ (Safioedin 1977) Balinese dÿÿu ‘king, chief (archaic for rÿtu)’; rÿtu ‘king; many persons commemorated in temples, semi-deified heroes, are given this title before their names’, hi rÿtu ‘pronoun, you (addressed to ksÿtriyÿs)’ (Barber 1979) Sasak dÿtu ‘prince’ (Goris 1938) Proto-Sangiric *datu ‘ruler, chief’ (Sneddon 1984) Sangir dÿtu ‘prince, king’, kÿ-rÿ-rÿtu-ÿN ‘time or place of ascent to kingship’ (Steller & Aebersold 1959) Bolaang Mongondow dÿtu ‘lord, prince, king’; ko-dÿtu-ÿn ‘be designated a prince’. In earlier times people hardly dared to use the word dÿtu, and generally called the prince ki tuÿN ‘the lord’ (Dunnebier 1951) Bare’e/Pamona dÿtu ‘(from Buginese) the title of the princes of Luwu’ who until 1906 were overseers of the Toraja people’ (Adriani 1928) Proto-South *datu ‘prince, chieftain’ [...] Further, Buginese, Makasarese dÿto? ‘old man, venerable’, also used as the title for early Islamic missionaries from Sumatra, hence probably < Malay dÿtok (Mills 1975) Makasarese dÿto? (from Malay) ‘title for distinguished persons, especially in the Malayicized groups of south Sulawesi’; dÿtu ‘Buginese princely title’; rÿtu ‘(including forms such as rÿtunnÿ with possessive suffix) old word for kÿrÿeN, prince’; kÿ-rÿtu-ÿN (old and poetic) ‘princely quality’ (Cense 1979)

The glosses of these forms present an interesting contrast with the related words in Philippine languages. Some, as those in South Sulawesi languages, and in the Toraja languages that they have influenced as a result of centuries of sociopolitical domination, show a clear association with the introduction of Islam by Malay-speaking missionaries. Many others, however, appear to hearken back to Hindu-Buddhist notions of the state and its rulers. This is particularly true in Javanese, and in those languages which fall within the sphere of historical Javanese cultural influence, as Sundanese, Madurese, Balinese, and (indirectly) Sasak. Sundanese and Makasarese rÿtu must be loans from Javanese, as *d > r is not a regular sound change in these languages. The fact that Makasarese rÿtu is given as an older form than dÿtu suggests that Javanese cultural influence reached this area in the Hindu-Buddhist period, and was followed at a later time by Islamic cultural influence brought by Malay-speaking missionaries from Sumatra or the Malay peninsula. Still other reflexes of *datu in western Indonesia refer either to respected elders within the kinship system (Malay, Acehnese, Gayo, Simalur) or to a shaman (Karo Batak, Toba Batak, 42 ROBERT BLUST

Dairi-Pakpak Batak, possibly Acehnese). In striking contrast to the adaptations of this Malay loanword in Philippine languages, where it is strongly associated with Islam and political leadership, the newly Islamized Dairi-Pakpak Batak of interior northern Sumatra associate the dÿtu (negatively) with the pre-Islamic role of the shaman, who is a traditional healer and communicator with the spirits of nature and of the dead. There are, then, three generally distinct blocks of glosses attached to reflexes of *datu in the Philippines and western Indonesia. The first is a title for Muslim petty rulers, especially prominent in the southern Philippines. The second is a title for princes or kings as conceived in the Hindu-Buddhist period of western Indonesian cultural history, a title that has survived more in literary than in practical contexts for some societies. Finally, there is a humbler sense, one connected with 1. kinship, 2. honored status as an elder or guardian of traditional law, and 3. shamanic practice. Each of these distinct blocks of glosses appears to represent an added layer of history brought about by the introduction of adventitious cultural influences: from a term of respect for a kinsman (marked morphologically by the presence of historical vocative suffixes in some languages, much as in other unambiguous kin terms), to a term of respect for the ruler of a Hindu-Buddhist state, to a term of respect for a titular Muslim official capable of commanding a following through inherited rank. In a sense much of the evidence we have looked at so far in trying to determine the original sense of *datu is tainted, since many of the reflexes of this word in western Indonesia and the Philippines are clear loanwords from Malay, while others are likely loanwords from Javanese. However, in a few languages that were relatively isolated from external influences until the twentieth century, as Gayo, the , or Simalur, the word appears to be native, and its sense is distinct from that in languages which historically fell within the so-called Malayosphere or Javanosphere—that is, the sphere of Malay or Javanese cultural domination. To see whether there is corroboratory evidence for this inference, we must consider the languages of eastern Indonesia and the Pacific, since to varying degrees these have been historically much freer, or in some cases even completely free, from Malay or Javanese linguistic and cultural influence. The following forms are known from languages of eastern Indonesia and the Pacific:

Savu rÿtu ‘prince’ (Wijngaarden 1896) Ngadha rÿtu ‘lord, king; noble, aristocrat’ (Arndt 1961) Kambera rÿtu ‘leader in religious ritual; priest’; rÿtu mÿrÿmb$ ÿ ‘priest of the prince, priest in service of the nobility’; mÿrÿpu rÿtu ‘oldest, now deified ancestors from whom superior lineages trace their ancestry’; mÿrÿpu = ‘mythical ancestors of a genealogical group’ (Onvlee 1984) Sika rÿtu ‘king’ (Meye 1964) Yamdena rÿtu ‘great man, king; guardian of the truth; God’ (Drabbe 1932b) Fordata rÿtu ‘king’ (Drabbe 1932a) Kei rÿt# ‘king, ruler’3 (Geurtjens 1921) Proto-Ambon *datu ‘king’ (Stresemann 1927)

3 Geurtjens (1921), who mistakenly associates Kei rÿt# with Malay (ultimately Sanskrit) rÿjÿ ‘king’, includes an extensive discussion of this word in connection with Kei myths of political origin. These hold that on Great Kei there originally was no rÿt# , but on Little Kei there were four. These were adopted on Great Kei, and the present population is still divided into groups that take their names from these ‘kings’. ON DATUS, ANCIENT AND MODERN 43

Asilulu lÿtu ‘king, village chief’; lÿtu kÿirupÿN ‘the ancient ruler of Asilulu before the descent to the coast, and the spirit which guards over the site of the old village’ (Collins, n.d.) Fijian (Bauan) rÿtu ‘honorific particle and title of rank before names of males who are chiefs’ (Capell 1968) Fijian (Wayan) rÿtý# ‘head of a clan, the title given to one who has been installed as a chief, possibly from Standard Fijian rÿtý# (Pawley & Sayaba, n.d.) Samoan lÿtý# ‘person in charge of an undertaking, e.g. a party of builders, or, in the case of women, a weaving bee’ (Milner 1966)

Next to these terms are several transparent Malay loanwords, as Tetun dÿto ‘the social class above the common people, the noble class’; dÿto fukun ‘important people’, or Lamaholot dÿtoâ ‘person in position of power or authority’. Before proceeding, some comments are necessary. Dempwolff (1938) included Fijian rÿtu, Samoan lÿtu as reflexes of *datu. Capell (1968) cites a similar word for Standard Fijian, but Pawley & Sayaba (n.d.) clearly mark both vowels in the Wayan (Western Fijian) form as long, and indicate that that same is true of Standard Fijian. Moreover, Milner (1966) gives Samoan lÿtý# also with both vowels long. Despite the close formal and semantic match of the Fijian and Samoan forms with those in insular Southeast Asia, then, these words, which apparently are cognate with one another, are best excluded from the comparison.4 This helps to resolve what was earlier a puzzling distribution, namely, widespread reflexes of this term in Southeast Asia (even if some of these are products of borrowing), but no attestation in any Oceanic language apart from the apparent cases in Fijian and Samoan. To return to our earlier question, what did PMP *datu mean? The answer is not at all obvious. Cognate forms that appear to be native are found both in Western Malayo- Polynesian (WMP) and in Central Malayo-Polynesian (CMP) languages, representing the two primary branches of Malayo-Polynesian.5 The meaning ‘king’ is given for a number of languages in each group (WMP: Aklanon, Sundanese, Javanese, Madurese, Balinese, Bolaang Mongondow, CMP: Ngadha, Sika, Yamdena, Fordata, Kei, Proto-Ambon, Asilulu), but this seems clearly to be secondary, and culturally inappropriate. If we instead substitute ‘chief’ the reconstruction immediately achieves greater plausibility, as the glosses ‘king’, ‘prince’, and ‘chief’ are widespread in WMP and CMP languages. However, this still does not incorporate two aspects of this comparison that merit further attention: 1) the clear indication in a number of WMP languages, that reflexes of *datu are associated either with senior kinsmen, or with ancestral kin, and 2) the relatively uncommon, but recurrent theme that reflexes of *datu refer to the role of a shaman or perhaps a lineage priest. In the hope of collecting more refined or elaborated glosses for some of these key meanings we need to turn to the literature of cultural anthropology. Vergouwen (1964) provides by far the most detailed available account of the social organization, religion, and customary law of the Toba Batak of northern Sumatra, who are

4 As a matter of method, it should be noted that striking sound-meaning associations which arise by chance are found in languages that have no known genetic relationship, as with Swahili kÿkÿ ‘elder brother’, Malay kÿkÿ ‘elder sibling’, or Yupik mÿnik ‘bird’s egg’, Gani mÿnik ‘bird’. It is sometimes overlooked, but important to keep in mind, that just because languages are genetically related does not mean that they cannot also share some striking sound-meaning associations that have arisen purely through convergence. 5 As noted elsewhere, WMP may not be a valid genetic unit. However, this is irrelevant in the present context, since in either case cognate forms that are found both in CMP languages and in one or more languages that have been assigned to a WMP group support a reconstructional inference that can be attributed to PMP. 44 ROBERT BLUST organized genealogically and politically into mÿrgÿ (patrilineal descent groups, of which the largest is the clan). He notes (1964:19) that the term dÿtu defines a class of people who function as priest and medicine-man in one. Over most of this region, Vergouwen writes: [...] the dÿtu have contributed much to the preservation of historical stories and the knowledge of the relationships within mÿrgÿ and tribal groups. In the Silahisabungan group of North-East Samosir the keeper of old of such knowledge was the boru silÿon, the representative of the oldest branch of the oldest in-dwelling mÿrgÿ, who directed the territory’s communal sacrifices as well as the collection of the contributions to them and the dividing up of the sacrificial animal, all according to the mÿrgÿ branches that inhabit the territory. The dÿtu is thus a lineage official who is responsible for the preservation of genealogical knowledge and the performance of ritual functions at major kin-related ceremonies. A strikingly similar function is described by van Wouden (1968:25–30) for the rÿtu among the Kambera of eastern Sumba: On Sumbÿ a district is always occupied by a number of named patrilineal clans called kÿbisu. In central and eastern Sumba the clan possesses its own territory, and the head of the clans is at the same time head of a village. It is his task to organise feasts, and to take care of the objects connected with the cult of the mÿrÿpu (ancestors) of the clan [...] At the head of the tribes (i.e. districts) stand rulers called mÿrÿmbÿ. The clan to which the ruler himself belongs is the most important kÿbisu in the district. We have seen how in the western part of the island, where a division of the tribe into halves is a phenomenon of general occurrence, the head of one of the halves has completely outstripped his counterpart in the possession of real power, whereas the latter nevertheless continues to be recognized as the older and the more eminent of the two. The dual organization is often accompanied by a differentiation of spiritual and worldly power. In central and eastern Sumba also this division of labour has been effected, but in an entirely different way. Everywhere in this region there are two categories of holders of power: the guardians of the relics of the mÿrÿpu, and the rulers. The task of the former, the rÿtu, consists chiefly in the performance of priestly functions; they also decide questions of inheritance and relationship, and all decisions by the ruler require their sanction. Furthermore, they are actually the mÿngu tÿnÿh (owners of the land), and they thus are invested with the dignity which is usually denoted by the Malay term tuÿntÿnÿh [...] Ruler and rÿtu are never united in one individual [...] in most districts there are four rÿtu, who are descendants of the oldest kÿbisu, those who in olden times first laid claim to the land. Evidently, therefore, the rÿtu is to be seen as having been originally the head of a clan. The office of rÿtu has become an entirely spiritual one, providing few material advantages [...] In western Sumba the ruler is at the same time the most important rÿtu. Van Wouden (1968) is a particularly valuable source of information on the role of the *datu in eastern Indonesia, as he has summarized much of the older and widely scattered literature in Dutch on traditional culture in this area. Citing the early twentieth-century Dutch colonial official O. D. Tauern, he notes (1968:71) that among the Wemale of Seram in the central Moluccas the lÿtu is ‘Lord of the Land’, a function that is widespread in eastern Indonesia. His primary responsibility is to maintain customary law, and decide matters of land tenure. According to one early Dutch observer the lÿtu nusÿ among the Wemale is a priest, and in some areas is leader of the youths at their initiation. There is thus a strong general agreement between ON DATUS, ANCIENT AND MODERN 45 the function of the dÿtu among the Toba Batak of northern Sumatra, the rÿtu among the Kambera of eastern Sumba, and the lÿtu among the Wemale of Seram. This is not to say that all reflexes of *datu describe an official who occupies the same or a closely analogous social role. Van Wouden (1968:67–70), for example, gives a somewhat different account of the function of the lÿtu for the Manusela, who live east of the Wemale in Seram. This group, he adds: [...] comprises four tribes, niniÿ niÿ or ÿmÿni, each divided into a number of patrilineal clans, uku, ibÿ, ifÿ or ifÿn, and a number of hamlets, lohoki [...] The ifÿn Hapisoa is also called IfÿnLÿtu, because the lÿtu, the head of the ÿmÿni, must always come from this ifÿn. The other ifÿn are therefore referred to as ifÿn rÿhekÿwÿsÿ, those without office. It is also said that in the time before the choice of the first lÿtu, there was already a distinction between noble and non-noble [...] The lÿtu governs the ÿmÿni as a whole, and his own lohoki in particular. He is in supreme command of any warlike undertaking. The Manusela lÿtu is further said to form, in connection with the heads of the lohoki and various other officials, a council which hears legal cases, and his office is normally passed from father to eldest son, but is also said to be elective within the ifÿn. Here the lÿtu appears to have a role that includes much more political power than is the case in the other examples previously reported, where the *datu tends to be opposed to the political leader as a complementary official best described as a ‘priest’ (or in modern terms, perhaps a role encompassing both that of a priest and a legal scholar). Finally, a function reminiscent of this is recorded in connection with a dÿtu among the Southern Toraja of central Sulawesi: The largest kin group is the ramage, i.e., an ancestor-oriented, ambilineal group. These pÿ’rÿpuÿn trace their descent to an ancestor who may have descended from heaven onto a mountain some twenty generations ago. The ancestor or his son founded a tongkonÿn, a house, in which the important ritual belongings of the family are stored and to which a title may be attached. As time went on new tongkonÿn were founded. Because of the ramage system the tongkonÿn are ordered hierarchically, and a person’s status is fixed by the tongkonan to which he belongs [...] As the tongkonÿn are ordered hierarchically, so are the titleholders. Thus in donÿlu the sokkong bÿju or dÿtu muÿne (the Male Lord) is the most important title, in diongnÿlu the dÿtu bÿine (Female Lord). As a rule both titleholders are men, though a woman of noble birth may also carry a title. (Nooy-Palm 1972:134–135) These more detailed descriptions go some way toward bridging the gap between such glosses as ‘king, chief’ and ‘priest, shaman’, but they still leave something to be desired. In consideration of the rather wide range of meanings to which reflexes of *datu apply, it might be thought that any attempt to ascertain the original sense is futile. Note, however, that if *datu did not originally refer to a lineage functionary, it is puzzling that it has come to have such a kin-based definition among at least the Simalur of northern Sumatra, the Southern Toraja of central Sulawesi, and various groups in the Lesser Sunda and Moluccan islands of eastern Indonesia. If *datu did refer to a lineage functionary, on the other hand, its transformation to a territorially based functionary could have been precipitated (in eastern Indonesia, at any rate), by the impact of European colonial administration, with its general policy of lumping together formerly distinct villages to facilitate census-taking, peace-keeping, and the like. In any case, the comparative linguistic evidence suggests that least four components of meaning must have entered into the definition of *datu. These are: 46 ROBERT BLUST

1) political leader, chief 2) priest, custodian, and administrator of customary law and genealogical knowledge (hence traditional scholar) 3) ancestor, elder, head of a kin group 4) noble

Before proceeding, another cognate set should be considered briefly, as it is relevant to determining the meaning of *datu. In a number of the languages of the central and southern Philippines, Sulawesi and Borneo, and Malay and Old Javanese reflexes of *balian appear with the meaning ‘shaman, medium, healer, seer’. Although reflexes of this form are unknown outside this region, it must be of considerable antiquity. Various sources mention that the *balian can be either male or female. In the case of males, the office was commonly associated with hermaphroditism or transvestitism (cf. Wilkinson 1959 sub beliÿn, Mills 1975 sub *bali+an). Unlike reflexes of *datu, reflexes of *balian never allude to political leadership, kinship, or nobility. It seems safe to conclude then, that at least as early as Proto-Western Malayo-Polynesian (PWMP) times there was an office of shaman or medium called *balian, which was filled by male or female persons recruited on a nonkinship basis.6 As appears to be almost universally the case, a prime qualification for office was almost certainly a physical, psychological, or behavioral abnormality—in other words, some tangible indication of extraordinary contact with the shadowy world of spirits. The office of *datu appears to have been filled by persons of a radically different type of character structure. Effective political leadership, or for that matter reliable information about genealogical or legal matters demands psychological and behavioral stability, characteristics completely foreign to the calling of the *balian. Similarly, although Wilkinson (1959) gives dÿto? perempuÿn in the meaning ‘grandmother’ and ‘wife of a principal chief’, and Nooy-Palm (1972) mentions that among the Southern Toraja a woman may bear the title dÿtu if she is of noble birth, in these cases and all others that have been examined in the ethnographic literature, it seems clear that *datu applied exclusively, or almost exclusively to males, even when—as among the Southern Toraja—the titleholder bears the appellation ‘Female Lord’.7 As a priest, the functions of the *datu must have complemented those of the *balian. The *balian can best be characterized as a shaman, an individual privy to the capricious ways of the supernatural world—that chaotic domain beyond the bounds of the structured world of kinship and community. By all indications, the *datu, on the other hand, was custodian of this very structured world of things predictable and customary. He was the indigenous genealogist and legal scholar, and where he dealt with the world of intangible beings it was through the performance of public sacrifice and ritual rather than of private communion with the spirits; in short, he dealt with the ancestors as the wellsprings of genealogy and the bestowers of customary law. While the *balian was a shaman, then, the *datu was a priest. Known comparative evidence provides no direct support for the inference that the *datu was a medical practitioner, since this function is reported for reflexes of *datu only in Sumatra.

6 Again, as noted in other publications, it is possible that WMP is not a valid linguistic subgroup, in which case many reconstructed terms that have been assigned to it may be more appropriately assigned to PMP. 7 As noted in Blust (1994:60–64), the terms ‘male’ and ‘female’ in the Southern Toraja terms for cross siblings (ÿnÿkdÿrÿ ‘sister of a man’, lit. ‘child’ + ‘virgin/maiden’, ÿnÿkmuÿne ‘brother of a woman’, lit. ‘child’ + ‘male’) almost certainly refer (or once referred) to symbolically ‘male’ (socially superior) and ‘female’ (socially subordinate) descent groups in a system of asymmetric alliance, or matrilateral cousin marriage. The terms “male dÿtu” and “female dÿtu” need to be considered in the same light. ON DATUS, ANCIENT AND MODERN 47

The notion of nobility appears in the definition of various reflexes of *datu. This dimension of the reconstructed meaning thus raises the question of a possible system of hereditary ranks in PMP society for which linguistic evidence (i.e., reconstructible terms for ‘noble’, ‘commoner’, and ‘slave’) is still incomplete. It is clear that *qadipen meant ‘slave’ at least as early as PWMP times, but etyma for ‘noble’ and ‘commoner’, unless they were *datu and *tau ‘human being’, are still to be sought. Nonetheless it is noteworthy that both the meanings ‘chief’ and ‘noble’ are widely associated with reflexes of *datu. If succession to high political office was entirely by merit in PMP times, there would be no obvious basis for this recurrent association. It thus appears likely that succession to positions of high authority was through noble lines. Although reflexes of *datu as a kinship term are not widespread (Malay, Gayo, Simalur, Barito languages of southeast Borneo), their appearance in languages that do not form a close-knit subgroup calls for explanation. If the office of *datu was not originally kin-based, these developments would be extremely puzzling. But if, as among the Toba Batak of northern Sumatra, Southern Toraja of central Sulawesi, and Kambera of eastern Sumba the *datu was a lineage or clan official (perhaps viewed as a descendant of the deified clan ancestors) the development of kinship meanings of the type ‘PaPaPa’ in Barito languages, or ‘grandfather’ in Malay could have developed by loss of the functional component of the original meaning, and elaboration of the kin-based component. There is another type of evidence which also suggests that the office of *datu was attached to a descent group. As noted by Nooy-Palm (1972), at the great buÿ?-kÿsÿlle? feast, the village is divided into donÿlu ‘upper’ and dioNÿlu ‘lower’ halves, each containing several prominent, hierarchically ordered lineages (toNkonÿn = ‘house’), which are associated with a title. The most important titles in the two halves are as follows.

donÿlu ‘upper’ dioNÿlu ‘lower’ sokkoN bÿju or dÿtu muÿne ‘male lord’ dÿtu bÿine ‘female lord’

Similarly, van Wouden (1968:152) notes that in collections of songs from Nusa Laut (Ambon) there is a clearly stated “opposition between the lÿtu of the upper side and the lÿtu of the lower side, who—to judge by their titles—are associated respectively with a wÿringin [banyan] and the water [...]” There is no obvious way in which this common terminological distinction could be a product of convergence. Borrowing likewise would appear to be out of the question as the Southern Toraja are, and presumably have long been, located in the mountainous interior of central Sulawesi, separated by more than eight hundred kilometers of sea and land from the people on the tiny island of Nusa Laut in the central Moluccas. By the normal canons of the Comparative Method, then, it would appear to be necessary to attribute expressions meaning ‘dÿtu of the upper half’ and ‘dÿtu of the lower half’ to Proto-Malayo-Polynesian. We have come a long way from Maranao dÿto?, where we began, and perhaps it is time to summarize what we have found. This matter was given cursory treatment in Blust (1980), where PMP *datu was glossed ‘clan priest’. As in any undertaking of this kind, there are gaps in the evidence, and frustrating uncertainties in certain areas. Nonetheless, the evidence we have strongly suggests that PMP *datu referred to a priest charged with the guardianship of the sacred paraphernalia of his lineage and of customary law, with the conduct of public rituals and ceremonies, and with the preservation of genealogical knowledge, among other things. This role was filled by a senior male of noble ancestry, and probably was transmitted from father to (perhaps eldest) son. It was clearly distinguished from that of the shaman, who was either male or female, and often a hermaphrodite or transvestite. The most difficult aspects of the cognate sets supporting *datu are their frequent reference to worldly authority, 48 ROBERT BLUST and the recurrent use of this word as a kin term for a supra-parental generation. As van Wouden noted, in some of the societies of eastern Indonesia there is a clear separation between the roles associated with secular authority and with sacred or scholarly authority. However, in other groups the two types of authority seem less clearly marked. Needham (1973:124) makes the same point with regard to many traditional societies, noting what he calls “the dual nature of sovereignty” in which one figure or class of individuals wields worldly power and its alter ego wields other-worldly power, although the two may be commingled in particular cases. If PMP *datu was a priest in a lineage-based society with inherited rank, we must ask what term designated the secular authority, or chief, and here we are at a loss, as no widely distributed linguistic comparison is known for this meaning other than *datu itself. Whether the *datu combined both roles as a priest-king, or was a priestly official who did not intrude upon secular authority remains an open question. In addition, the PMP term *datu may have referred both to a living person and to the ancestral line from which he came (hence its recurrent use as a kin term). Needless to say, the attempt to assign a well motivated gloss to this one term has much wider implications for the structure of PMP society, implying minimally that it had ancestor-oriented corporate kin groups and hereditary distinctions of rank, features that persisted in some attested societies long after they had been significantly transformed by external contacts.

References

Adriani, N. 1928. Bare’e-Nederlandsch woordenboek. Leiden: Brill. Arndt, Paul. 1961. Wörterbuch der Ngadhasprache (Studia Instituti Anthropos 15.) Posieux (Fribourg), Switzerland: Anthropos Institute. Awed, Silin A., Lillian B. Underwood, & Vivian M. Van Wynen. 2004. Tboli-English dictionary. Manila: Summer Institute of Linguistics. Barber, C. Clyde. 1979. Dictionary of Balinese-English, 2 vols. (Aberdeen University Library Occasional Publications 2). Aberdeen: University of Aberdeen. Behrens, Dietlinde. 2002. Yakan-English dictionary (Linguistic Society of the Philippines Special Monograph 40.2). Manila: Linguistic Society of the Philippines. Benton, Richard A. 1971. Pangasinan dictionary. Honolulu: University of Hawaii Press. Bergaño, Diego. 1732/1860. Vocabulario de la lengua pampanga en romance. Manila: Ramirez y Giraudier. Blust, Robert. 1979. Proto-Western Malayo-Polynesian vocatives. Bijdragen tot de Taal-, Land- en Volkenkunde 135. 205–251. Blust, Robert. 1980. Early Austronesian social organization: The evidence of language. Current Anthropology 21(2). 205–247 (including comments and reply). Blust, Robert. 1994. Austronesian sibling terms and culture history. In A. K. Pawley & M. D. Ross (eds.), Austronesian terminologies: Continuity and change (Pacific Linguistics C–127). Canberra: Department of Linguistics, Research School of Pacific and Asian Studies, Australian National University. 31–72. Capell, A. 1941/1968. A new Fijian dictionary. Suva: Government Printer (1st edn., Sydney: Australasian Medical Publishing). Cense, A. A. 1979. Makassaars-Nederlands woordenboek. The Hague: Nijhoff. ON DATUS, ANCIENT AND MODERN 49

Cole, Fay-Cooper. 1913. The wild tribes of the Davao district, Mindanao. Publication 170. (Field Museum of Natural History Anthropological Series 12.2). Chicago: Field Museum of Natural History. 49–203. Collins, James T. [N. d.] Asilulu dictionary. Typescript. Collins, Millard A., Virginia R. Collins, & Sulfilix A. Hashim. 2001. Mapun-English dictionary. Manila: Summer Institute of Linguistics Philippines. Coolsma, S. 1884/1930. Soendaneesch-Hollandsch woordenboek, 2nd edn. Leiden: Sijthoff. Dempwolff, Otto. 1938. Austronesisches Wörterverzeichnis: Vergleichende Lautlehre des austronesischen Wortschatzes 3 (Beihefte zur Zeitschrift für Eingeborenen-Sprachen 19). Berlin: Dietrich Reimer. Drabbe, P. 1932a. Woordenboek der Fordaatsche taal (Verhandelingen van het Bataviaasch Genootschap van Kunsten en Wetenschappen, 71.2). Bandung: A. C. Nix. Drabbe, P. 1932b. Woordenboek der Jamdeensche taal (Verhandelingen van het Bataviaasch Genootschap van Kunsten en Wetenschappen 71.3). Bandung: A. C. Nix. Dunnebier, W. 1951. Bolaang Mongondowsch-Nederlandsch woordenboek. The Hague: Nijhoff. Elkins, Richard E. 1968. Manobo-English dictionary (Oceanic Linguistics Special Publications 3). Honolulu: University of Hawaii Press. Forman, Michael L. 1971. Kapampangan dictionary. Honolulu: University of Hawaii Press. Geurtjens, H. 1921. Woordenlijst der Keieesche taal (Verhandelingen van het Bataviaasch Genootschap van Kunsten en Wetenschappen 63.3). The Hague: Nijhoff. Goris, R. 1938. Beknopt Sasaksch-Nederlandsch woordenboek. Singaradja, Bali: Kirtya Liefrinck-van der Tuuk. Hassan, Irene U., Seymour A. Ashley, & Mary L. Ashley. 1975/1994. Tausug-English dictionary: Kabtangan iban maana, 2nd edn. (Sulu Studies 6). Jolo, Sulu: Notre Dame of Sulu College; Manila: Summer Institute of Linguistics (1st edn., Sulu Studies 4, Jolo, Sulu: Coordinated Investigation of Sulu Culture; Manila: Summer Institute of Linguistics). Horne, Elinor Clarke. 1974. Javanese-English dictionary. New Haven: Yale University Press. Hudson, Alfred B. 1967. The Barito isolects of Borneo: A classification based on comparative reconstruction and lexicostatistics (Data Paper 68). Ithaca: Department of Asian Studies, Cornell University. Kähler, Hans. 1961. Simalur-Deutsches Wörterbuch mit Deutsch-Simaluresischem Wörterverzeichnis. Berlin: Dietrich Reimer. Kreemer, J. 1931. Atjèhsch handwoordenboek. Leiden: Brill. Laktaw, Pedro Serrano. 1914/1965. Diccionario tagálog-hispano: Segunda parte. Manila: La Opinión (2nd edn., Madrid: Ediciones Cultura Hispánica). LeBar, Frank M. (ed. and compiler). 1975. Philippines and Formosa (Ethnic Groups of Insular Southeast Asia 2). New Haven: Human Relations Area Files Press. 50 ROBERT BLUST

Manik, Tindi Radja. 1977. Kamus Bahasa Dairi-Pakpak-Indonesia. Jakarta: Pusat Pembinaan dan Pengembangan Bahasa, Departemen Pendidikan dan Kebudayaan. McKaughan, Howard P., & Batua A. Macaraya. 1967/1996. A Maranao dictionary. Honolulu: University of Hawaii Press (2nd edn., Manila: De La Salle University Press and Summer Institute of Linguistics). Melalatoa, M. J., Harimurti Kridalaksana, Ahmad Banta, Asiah Abubakar Melalatoa, & Abu Daud Benerselam. 1985. Kamus Bahasa Gayo-Indonesia. Jakarta: Pusat Pembinaan dan Pengembangan Bahasa, Departemen Pendidikan dan Kebudayaan. Meye, Frans. 1964. Sikaneesch-Hollandsch woordenboek. Unpublished typescript, Ruteng, Flores. Mills, Roger F. 1975/1976. Proto South Sulawesi and Proto Austronesian phonology, 2 vols. Ann Arbor: University Microfilms International (published version of Roger Frederick Mills, Proto South Sulawesi and Proto Austronesian phonology, Ann Arbor: University of Michigan PhD dissertation). Milner, G. B. 1966. Samoan dictionary. London: Oxford University Press. Mintz, Malcolm W. 1971/2004. Bikol dictionary: Diksionáryong Bíkol. Perth: Indonesian/Malay Texts and Resources (1st edn., Bikol dictionary, Honolulu: University of Hawaii Press). Needham, Rodney. 1973. The left hand of the Mugwe. In Rodney Needham (ed.), Right and left: Essays on dual symbolic classification. Chicago: University of Chicago Press. 109–127. Neumann, J. H. 1951. Karo-Bataks-Nederlands woordenboek. Medan, Sumatra: Varekamp. Nooy-Palm, C. H. M. 1972. Southern Toraja. In Frank M. LeBar (ed. and compiler), Indonesia, Andaman Islands, and Madagascar (Ethnic Groups of Insular Southeast Asia 1). New Haven: Human Relations Area Files Press. 133–136. Onvlee, L. 1984. Kamberaas (Oost-Soembaas)-Nederlands woordenboek. The Hague: Nijhoff. Panganiban, Jose Villa. 1966. Talahuluganang Pilipino-Ingles. Manila: Kawanihan ng Palimbagan. Pawley, Andrew K., & Timoci Sayaba. [2003]. Words of Waya: A dictionary of the Wayan dialect of the Western . Unpublished manuscript. Post, Ursula, & Mary Jane Gardner. 1992. Binukid dictionary (Studies in Philippine Linguistics 9.2). Manila: Linguistic Society of the Philippines and SIL Philippines. Rubino, Carl R. Galvez. 2000. Ilocano dictionary and grammar. Honolulu: University of Hawai‘i Press. Safioedin, Asis. 1977. Kamus Bahasa Madura-Indonesia. Jakarta: Pusat Pembinaan dan Pengembangan Bahasa, Departemen Pendidikan dan Kebudayaan. Salas Reyes, Vicente, R. David Paul Zorc, & Nicolas L. Prado. 1969. Dictionary (of root words and derivations), Aklanon to English (A Study of the Aklanon Dialect 2). Kalibo, Aklan: Public Domain. Schlegel, Stuart A. 1971. Tiruray-English lexicon (University of California Publications in Linguistics 67). Berkeley: University of California Press. ON DATUS, ANCIENT AND MODERN 51

Sneddon, J. N. 1984. Proto-Sangiric and the Sangiric languages (Pacific Linguistics B–91). Canberra: Department of Linguistics, Research School of Pacific Studies, Australian National University. Steller, K. G. F., & W. E. Aebersold. 1959. Sangirees-Nederlands woordenboek met Sangirees register. The Hague: Nijhoff. Stresemann, Erwin. 1927. Die Lauterscheinungen in den ambonischen Sprachen (Beiheft zur Zeitschrift für Eingeborenen-Sprachen 10). Berlin: Reimer. Svelmoe, Gordon, & Thelma Svelmoe. 1990. Mansaka dictionary (Language Data, Asia-Pacific Series 16). Dallas: Summer Institute of Linguistics. Toorn, J. L. van der. 1891. Minangkabausch-Maleisch-Nederlandsch woordenboek. The Hague: Nijhoff. Vergouwen, J. C., & Jeune Scott-Kemball (transl.). 1964. The social organization and customary law of the Toba-Batak of northern Sumatra (Koninklijk Instituut voor Taal-, Land- en Volkenkunde Translation Series 7). The Hague: Nijhoff. Warneck, Joh. 1906/1977. Toba-Bataks-Deutsches Wörterbuch. The Hague: Nijhoff (1st edn., Tobabataksch-deutsches Wörterbuch, Batavia: Landsdrukkerij). Wijngaarden, J. K. 1896. Sawuneesche woordenlijst. The Hague: Nijhoff. Wilkinson, R. J. 1959. A Malay-English dictionary (romanised), 2 parts. London: Macmillan. Wolfenden, Elmer P. 2001. A Masbatenyo-English dictionary (Linguistic Society of the Philippines Special Monograph 38). Manila: Linguistic Society of the Philippines. Wolff, John U. 1972. A dictionary of Cebuano Visayan, 2 vols. (Data Paper 87). Ithaca, NY: Southeast Asia Program, Cornell University. Wouden, F. A. E. van, & Rodney Needham (transl.). 1935/1968. Types of social structure in eastern Indonesia (Koninklijk Instituut voor Taal-, Land- en Volkenkunde Translation Series 11). The Hague: Nijhoff (1st edn., Sociale structuurtypen in de groote Oost, Leiden: J. Ginsberg). Zoetmulder, P. J. 1982. Old Javanese-English dictionary, 2 vols. The Hague: Nijhoff.